US Media: Zerologon GLM-5.2 Ranks Alongside Mythos in Vulnerability Discovery

US0%
XVG0%
OPUS0%
MYTH0%
GLM0%

BlockBeats News, June 29th, According to a report from the US tech media The Verge and The Wall Street Journal, the domestic heavyweight AI company Silkbrain (02513.HK) recently released its open-weight model GLM-5.2. Although GLM still lags behind Anthropic and OpenAI models in other more general tasks, some researchers claim that this model can rival Anthropic's Mythos model in certain vulnerability detection and cybersecurity scenarios.

According to the cybersecurity company Semgrep, in certain benchmark tests, GLM-5.2 has outperformed Anthropic's Claude Opus 4.8 model released in May. Researchers point out that with further instructions, Opus 4.8 and GLM-5.2 can match Mythos in vulnerability discovery ability.

However, since GLM is an open-weight model that anyone can download and run on readily available hardware, although this gives it great flexibility and allows advanced users deep access, it may also increase the risk of being exploited by malicious actors.

---------------------------------
Click on the original text link below to join the BlockBeats · Feishu AI News Channel, which monitors global AI hotspots and news 24/7.

ソース：BlockBeats

免責事項：現在のコンテンツは第三者の視点に基づくもの、または第三者の視点からAIが直接翻訳したものです。CoinExはコンテンツの信頼性、正確性、独創性を保証するものではなく、CoinExからの投資アドバイスを構成するものではありません。暗号資産の価格変動は急激に変動します。潜在的なリスクにご注意ください。